一种基于UKF的矿井智能体相对导航算法

A Relative Navigation Algorithm for Mine Agents Based on Unscented Kalman Filter

以对偶四元数为理论基础,引入非线性滤波UKF算法,构建矿井多智体视觉自主相对导航算法DQUKF,并建立适应矿井智能体外形特征及其工作环境的算法状态方程及观测方程,可对矿井智能体的相对位置、姿态进行动态估计及仿真。为验证和评价该算法的有效性,设计地面实物模拟实验场,结果表明,该算法能够提供较为精确的相对位姿信息,结论与仿真结果一致。

To dynamically estimate and simulate the relative position and attitude of the mine intelligent body,we take the dual quaternion as the theoretical basis,introduce the nonlinear filtering UKF algorithm,study and construct the mine multi-intelligent body visual autonomous relative navigation algorithm DQUKF,and establish the algorithm state equation and observation equation adapted to the shape characteristics of the mine intelligent body and its working environment.To verify and evaluate the effectiveness of this paper’s algorithm,we design a ground physical simulation experimental field,and the results show that this paper’s method can provide more accurate relative position and attitude information,and the conclusions are consistent with the simulation results.